Hybrid method based on the inverse wavelet transform and hopfield network to reconstruct the transformed data

Issue title: Special Volume: Proceedings of the eleventh International Symposium on Applied Electromagnetics and Mechanics ISEM-Versailles ISEM 03

Article type: Research Article

Authors: Nouicer, Abdelmadjid^{a; *} | Latreche, M.E.H.^a | Feliachi, Mouloud^b

Affiliations: [a] Département d'électrotechnique, Université Mentouri, route Ain Elbey,Constantine (25000), Algéerie | [b] LRTI /GE44Laboratory, Saint Nazaire, France

Correspondence: [*] Corresponding author. E-mail: nouicer_abdelmadjid@yahoo .fr

Abstract: Conventional reconstruction of the compressed data based on the inverse wavelet transform (IWT) does not give good results (too much loss of information) because of the neglected detail part especially for the high levels decomposition. To deal with this problem, a new approach for the reconstruction of the compressed data, based on the inverse wavelet transform and the Hopfield neural networks is presented in this paper.

Keywords: electromagnetic problems, FEM, hopfield network, wavelet transform

DOI: 10.3233/JAE-2004-637

Journal: International Journal of Applied Electromagnetics and Mechanics, vol. 19, no. 1-4, pp. 607-611, 2004